Soft robotic systems are highly susceptible to sensor noise arising from hardware imperfections, environmental disturbances, and the intrinsic compliance of soft materials. These noisy measurements can obscure essential state information and degrade performance in both perception and control tasks. In this paper, we introduce a Contrastive Denoising Variational Recurrent Neural Network (CD-VRNN) designed to address this challenge. The proposed model learns to decompose time-series data into clean signal and noise pathways, improving state estimation accuracy for soft robotic platforms. By incorporating a contrastive objective that enforces a separation between signal-related and noise-related latent representations, the proposed CD-VRNN more effectively isolates noise while retaining critical temporal features. In addition, we incorporate conditional flow-based priors for expressive, state-dependent distributions and skip-connected decoders that preserve subtle signal variations. Experimental results on a pneumatic soft robot and multiple public time-series datasets show that CD-VRNN consistently outperforms existing approaches for denoising and downstream task performance, demonstrating the model’s robustness and strong generalization capacity.

Soft robotic systems heavily depend on accurate sensor data for perception and control; however, this data is often corrupted by missing observations, due to partial sensor coverage, communication failures, or occlusions and noisy measurements stemming from hardware imperfections, environmental disturbances, and the intrinsic compliance of soft materials. Such corruption can obscure critical state information, causing unreliable modeling of soft robotics and degrading control accuracy. To address these challenges, we propose a Contrastive Dual-Latent Autoencoder (CDLAE) that jointly handles missing and noisy data in a single end-to-end framework. Our approach leverages an attention based autoencoder architecture with dual latent pathways, where one focuses on capturing the underlying clean signals while the other isolates noise-related components. A contrastive loss encourages strong separation between these pathways, enhancing the model’s ability to filter noise while reconstructing missing values. Additionally, the autoencoder is trained jointly with a downstream predictive network, ensuring that signal imputation is optimized with respect to the ultimate control task. Experimental evaluations on a pneumatic soft robot platform and multiple public time-series datasets demonstrate that CDLAE consistently outperforms existing methods in handling corrupted data, offering robust, high-fidelity reconstructions that significantly improve soft robot perception and control in real-world conditions.

Most works on joint state and unknown input (UI) estimation require the assumption that the UIs are linear; this is potentially restrictive as it does not hold in many intelligent autonomous systems. To overcome this restriction and circumvent the need to linearize the system, we propose a derivative-free UI sigma-point Kalman filter (SPKF-nUI), where the SPKF is interconnected with a general nonlinear UI estimator that can be implemented via nonlinear optimization and data-driven approaches. The nonlinear UI estimator uses the posterior state estimate, which is less susceptible to state prediction error. In addition, we introduce a joint sigma-point transformation scheme to incorporate both the state and UI uncertainties in the estimation of SPKF-nUI. An in-depth stochastic stability analysis proves that the proposed SPKF-nUI yields exponentially converging estimation error bounds under reasonable assumptions. Finally, two case studies are carried out on a simulation-based rigid robot and a physical soft robot, i.e., the robots made of soft materials with complex dynamics, to validate the effectiveness of the proposed filter on nonlinear dynamic systems. Our results demonstrate that the proposed SPKF-nUI achieves the lowest state and UI estimation errors when compared to the existing nonlinear state-UI filters.

Recently, data-driven models such as deep neural networks have shown to be promising tools for modelling and state inference in soft robots. However, voluminous amounts of data are necessary for deep models to perform effectively, which requires exhaustive and quality data collection, particularly of state labels. Consequently, obtaining labelled state data for soft robotic systems is challenged for various reasons, including difficulty in the sensorization of soft robots and the inconvenience of collecting data in unstructured environments. To address this challenge, in this paper, we propose a semi-supervised sequential variational Bayes (DSVB) framework for transfer learning and state inference in soft robots with missing state labels on certain robot configurations. Considering that soft robots may exhibit distinct dynamics under different robot configurations, a feature space transfer strategy is also incorporated to promote the adaptation of latent features across multiple configurations. Unlike existing transfer learning approaches, our proposed DSVB employs a recurrent neural network to model the nonlinear dynamics and temporal coherence in soft robot data. The proposed framework is validated on multiple setup configurations of a pneumatic-based soft robot finger. Experimental results on four transfer scenarios demonstrate that DSVB performs effective transfer learning and accurate state inference amidst missing state labels.

Soft sensors are widely used in many industrial systems to monitor key variables that are difficult to measure, using measurements from other available physical sensors. Because physical sensors are susceptible to faults, it is crucial for soft sensor models to be robust against them. Recently, deep learning has shown promising results in developing data-driven soft sensors for various applications. However, existing learning-based soft sensors are still vulnerable to sensor faults, which could deteriorate the performance of the models. In this article, we propose a deep learning-based modeling framework for developing soft sensor models that are robust to sensor faults. Due to the difficulty in obtaining datasets that cover all possible sensor fault characteristics, the proposed framework is developed to be zero-shot such that the model can be trained with only fault-free dataset without requiring any sensor fault patterns, thus greatly saving the time and resources needed to collect such data. Instead, adversarial examples are used as a proxy for faulty sensor inputs so that the model can learn to be adaptive through the proposed two-stage, uncertainty-aware recurrent neural network architecture. We demonstrate our approach to the TE benchmark process and a real industrial multiphase flow process and show that robustness is achieved as the accuracy does not degrade significantly when sensor faults are present during the model evaluation.

Bacterial chemotaxis in unicellular Escherichia coli, the simplest biological creature, enables it to perform effective searching behaviour even with a single sensor, achieved via a sequence of "tumbling" and "swimming" behaviours guided by gradient information. Recent studies show that suitable random walk strategies may guide the behaviour in the absence of gradient information. This article presents a novel and minimalistic biologically inspired search strategy inspired by bacterial chemotaxis and embodied intelligence concept: a concept stating that intelligent behaviour is a result of the interaction among the "brain," body morphology including the sensory sensitivity tuned by the morphology, and the environment. Specifically, we present bacterial chemotaxis inspired searching behaviour with and without gradient information based on biological fluctuation framework: a mathematical framework that explains how biological creatures utilize noises in their behaviour. Via extensive simulation of a single sensor mobile robot that searches for a moving target, we will demonstrate how the effectiveness of the search depends on the sensory sensitivity and the inherent random walk strategies produced by the brain of the robot, comprising Ballistic, Levy, Brownian, and Stationary search. The result demonstrates the importance of embodied intelligence even in a behaviour inspired by the simplest creature.

Morphological computation is a concept relevant to robots made of soft and elastic materials. It states that robot's rich dynamics can be exploited to generate desirable behaviors, which can be altered when their morphology is adapted accordingly. This paper presents a low-cost robot made of elastic curved beam driven by a motor, with morphological computation and adaptation ability. Simply by changing robot's shape and the rotating frequency of the motor that vibrates the robot's body, the robot is able to shift its behavior from showing a tendency to slide when it needs to perform tasks like going under confined space, to have more tendency to hop diagonally forward when the robot stands upright. It will also be shown that based on the proposed mechanism, the energy efficiency of the robot locomotion can be maximized.

The realization and utilization of multimodal locomotion to enable robots to accomplish useful tasks is a significantly challenging problem in robotics. Related to the challenge, it is crucial to notice that the locomotion dynamics of the robots is a result of interactions between a particular control structure and its body-environment dynamics. From this perspective, this paper presents a simple control structure known as Attractor Selection Mechanism that enables a robot to self organize its multiple locomotion modes for accomplishing a goal-directed locomotion task. Despite the simplicity, the approach enables the robot to automatically explore different body-environment dynamics and stabilize onto particular attractors which corresponds to locomotion modes relevant to accomplish the task. The robot used throughout the paper is a curved-beam hopping robot, which despite its simple actuation method, possesses rich and complex body-environment dynamics.

Robots composed of soft materials can achieve high deformability and conformity with unstructured and dynamic environments due to their body mechanics. However, it is challenging to gather desirable information about these robots' behaviors as conventional sensory systems are not designed to detect infinite degrees of freedom on continuum bodies. This paper presents a technical method to sensorize a soft elastic body by using conductive thermoplastic elastomer (CTPE) based strain gauge sensors. Due to its thermoplastic nature, CTPE based sensors can be fabricated in flexible sizes and shapes and integrated into the soft elastic bodies. We analyze soft elastic deformations and extract strain vectors to find finite unique regions on continuum surfaces. Then we use these regions to design the morphologies of our fiber shaped CTPE sensors to discriminate soft robotic behaviors. To demonstrate our approach and show how the programmable sensor morphologies discriminate different motion patterns, we have built a soft elastic prismatic silicone (E = 1.31MPa) block and designed two sensors to discriminate serpentine and twisting patterns.

Biological creatures, some of them are very simple, seem to perform efficient search strategy. Recent researches show that noise in their internal mechanism may have an important role to manage this behavior. This paper focuses on realizing a simple, noise utilizing, mathematical framework that enables a mobile robot to perform random search adaptively and efficiently under changing target density. Our approach is to model and implement bacterial movement based on a recent perspective of noise utilizing mechanism in living beings: biological fluctuation. As a result, the robot will adaptively switch its random search pattern between Levy walk and Brownian walk to increase the search efficiency in a patchy environment where the target density naturally alternates.

Many biologically inspired approaches have been investigated in relation with researches on mobile robot(s) that can effectively locate targets that induce gradient information. Here, we concentrate on realizing an adaptive searching behavior in mobile robot that is simple, yet effective with and without sensing the gradient information. We are interested in two searching behaviors found in biological creatures: bacterial chemotaxis, probably the simplest yet effective gradient sources searching behavior found in living creatures; and Levy walk, specialized random walks with fractal movement trajectories that optimize random search for sparsely and randomly distributed target(s). Our approach is to implement and combine the two searching behaviors based on ¿yuragi¿, a Japanese word for biological fluctuation.